The present study assesses the effect of accounting the temporal variation of near-surface lapse rate in the conceptual, degree-day snowmelt runoff model simulations in a cold-desert region of Himalayas. The nearsurface lapse rate over Spiti basin shows seasonal variation during a year. The results obtained show that the inclusion of monthly variation of lapse rate in the hydrological modelling is able to capture the observed hydrograph more efficiently than when an annually constant value of lapse rate is employed. Based on our results and considering the available data, a monthly representation of near-surface lapse rates in the temperature index based models is recommended for Himalayan basins.